Tegra 4. Source: NVIDIA

Broadcom (UNKNOWN:BRCM.DL) is officially bailing on basebands, thanks largely to the intense R&D investments required that simply weren't paying off in the face of intense competition. Challenging top dog Qualcomm and low-cost rising star MediaTek was simply proving too hard. Does that mean that other players should be eyeing the exits?

Who's coming with me?
(NASDAQ:NVDA) hasn't had much luck in cracking the difficult baseband market. NVIDIA CEO Jen-Hsun Huang acknowledged long ago that the discrete baseband market was on the decline as integrated solutions took over the market. That's why Tegra 4i, which features integrated cellular connectivity, was such an important launch.

Tegra 4i-powered devices are expected to hit the market later this year, but Huang has acknowledged that Tegra 4i has not performed well overall. Even though Tegra 4i is on its way to market, the company has confirmed that it's not focusing anymore on developing processors with integrated cellular connectivity. Tegra 4i will be NVIDIA's first and last attempt at an integrated modem/applications processor.

Instead, it will shift focus back to stand-alone basebands -- the same market that Huang previously conceded was limited. Meanwhile, NVIDIA has hit a wall in the mainstream smartphone market, and the company is trying to pivot into other verticals like automotive. 

It's true that connected cars are on the rise, which will be an opportunity for baseband suppliers, but cars are a relatively low-unit market compared to something like smartphones and tablets. Global auto sales were around 83 million last year. There were over 1 billion smartphones and nearly 77 million tablets shipped last year.

Due to the R&D intensity of cellular basebands, vendors really need to ramp up volume in order to scale profitably. Otherwise they risk never recouping those investments. That's exactly what happened to Broadcom and it's the same path that NVIDIA is heading down.

Mobile money pit
NVIDIA's Icera basebands are included in the broader Tegra division's results, and that segment fell off a cliff last fiscal year.

Source: 10-K. Fiscal years ending January shown.

The company plunged $600 million into Tegra R&D in 2013. At the time, it was hoping those investments would pay off as it targeted smartphones, tablets, Windows RT devices, automotive, and other gaming devices. Of those five target markets, NVIDIA has lost most of its traction in three of them.

Just this week, NVIDIA presented at the Bank of America Merrill Lynch Global Technology Conference. CFO Colette Kress echoed the notion that NVIDIA is simply refocusing on visual computing (where it is an undisputed leader) at large, regardless of what form factor that implies.

It should be somewhat telling that Broadcom's decision to exit the baseband market was only shortly after its wireless division swung to an operating loss. Broadcom could see the writing on the wall and knew it was time to get out before losses widened. NVIDIA has still not recognized this.

In addition to Tegra's $268 million operating loss last fiscal year, the segment burned through an additional $61 million in operating losses. That's an awful lot for a company with a total market cap of $10.5 billion.

It's time to get out, NVIDIA.